텍스트 네트워크 분석을 통한 부산시 버스민원 패턴분석

Abstract

Advancements in big data analysis techniques have enabled a wide range of novel studies to be conducted in various academic fields. However, very little research has employed big data to examine civil complaints, in particular bus-related grievances, in relation to local government policies. Thus, the purpose of this study is to analyze such bus-related civil complaints, which account for the majority of the electronic civil complaints received by Busan City Hall. Utilizing 10,380 cases of electronic civil complaints from 2015 to 2017, the study chose to analyze them from a microscopic rather than a macroscopic viewpoint. The study extracted main keywords related to bus using keyword analysis and text mining. Then, it visualized closeness and correlation among keywords through text network analysis to determine the change and importance among the main keywords related to buses for three years. The continuous correlation analysis among keywords and main civil complaint keywords helped increase our understanding of bus-related complaints from various viewpoints. First, the results revealed that home-work commuting complaints by citizens were the most common. The related keywords were “increase in running buses,” “operating hours,” and “bus route extension,” which had the most correlated keywords. Second, the main civil complaints in 2015 and 2016 were “increases in bus routes and running buses” and “frequency of bus arrival interval” due to home-work commuting. Other issues that were reported included unfriendly bus drivers, the violation of regular operations, and reckless driving during bus service in 2017. Third, correlations among keywords, such as “increase in running buses”, “operating hours”, and “home-work commuting” could be analyzed together as civil complaints that require more buses and the adjustment of operating hours to resolve the issues during workers’ commutes. The results of this study are expected to be employed as preliminary data that could assist in the creation of countermeasures to identify areas of improvement, increase the responsiveness of administration, and reduce the number of civil complaints.